Just to clarify, I have one live instance, but no resources that are
managed by Helix. Is it possible that no participants meet the criteria
because there are no resources (which could be the case if the call is
interpreted to mean all participants that host resources that qualify
the criteria).
If this is the reason why the criteria is not met by participants, is
there a way to have a message delivered to every participant regardless
of the resources hosted by it?
Thanks,
Vinayak
On 10/7/13 10:30 PM, Zhen Zhang wrote:
Hi, could you try allParticipantsCriteria.setSelfExcluded(false), by
default selfExcluded is true and maybe you are sending from the only
liveInstance in the cluster, so no liveInstance matches the criteria.
Thanks,
Jason
On Mon, Oct 7, 2013 at 9:30 PM, Vinayak Borkar <[email protected]> wrote:
Hi Kishore,
As suspected, not setting resource and partition did not help. Any other
ideas?
Thanks,
Vinayak
On 10/7/13 7:04 PM, kishore g wrote:
Hi Vinayak,
Why is resource and partition empty string? if you want to send all live
instances irrespective what resource/partition they do not set it.
Criteria allParticipantsCriteria = new Criteria();
allParticipantsCriteria.****setInstanceName("%");
allParticipantsCriteria.****setRecipientInstanceType(****InstanceTyp
e.PARTICIPANT);
can you try this
Thanks,
Kishore G
On Mon, Oct 7, 2013 at 6:58 PM, Vinayak Borkar <[email protected]>
wrote:
Hi,
I need to send a message to all live instances in a Helix cluster. I set
the Criteria as follows:
Criteria allParticipantsCriteria = new Criteria();
allParticipantsCriteria.****setInstanceName("%");
allParticipantsCriteria.****setRecipientInstanceType(**
InstanceType.PARTICIPANT);
allParticipantsCriteria.****setResource("");
allParticipantsCriteria.****setPartition("");
However, the number of recipients turns out to be 0. I also tried
changing
the datasource in the criteria to be IDEALSTATES using the code below:
allParticipantsCriteria.****setDataSource(DataSource.****IDEALSTATES);
But no joy.
Please note that I connect to Helix using an administrator role. When
listing the live instances using the helix-admin client script I do see
one
live instance.
Any ideas on what could be going wrong?
Thanks,
Vinayak